Pavement Image Segmentation Based on FCM Algorithm Using Neighborhood Information

Xinsong Wang (Tongji University)
Guofeng Qin (Tongji University)

Indonesian Journal of Electrical Engineering and Computer Science
Vol 10, No 7: November 2012

Standard FCM algorithm takes the pixel gray-scale information into account only, while ignoring the spatial location of pixels, so the standard FCM algorithm is sensitive to noise. This paper present a pavement image segmentation algorithm based on FCM algorithm using neighborhood information. The presented algorithm introduces neighborhood information into membership function to improve the standard FCM algorithm. It can eliminate noise effectively and retain the boundary information. The experiments by synthetic images and real pavement images show that the presented algorithm in this paper performs more robust to noise than the standard FCM algorithm and retain the boundary information effectively. DOI: http://dx.doi.org/10.11591/telkomnika.v10i7.1551